i think it was from my floor feeding. The amount of pressure from the material on the floor seemed to be the main driver. I could never get it to lay consistently. I wasn't managing it well since it was the purge. I bought a beer key bucket hoping that would help it to loose coil, and I couldn't get it to work. I did run some real filament afterwards where I payed closer attention and I was getting a lot less variance- 1.68-1.72 for most of the run with some readings out of that which I think came from my bad gauging technique since an inch or so away was fine. I did hand roll it as a psuedo filawinder and that helped alot. Definitely see a Filawinder in my future.
I'm using the vertical set up and I do find that the beads don't feed perfectly. Right now I only have the throat on since I wanted to keep things simple and run a couple of test colors and materials. More feed above might help it feed better. 15 seconds of tapping every 10 grams of filament gets it to feed better. I do think after I do this that I hear some pellets perhaps 'grinding' on the hopper, perhaps between the screw and hopper. Can you over feed the screw?
I'm thinking about printing something that would attach to the screw on the flat section and as it turns have it cyclically prime a spring driver kicker that would tap the throat.
That's why I asked about over feeding it since that would affect the design- 3-4 whacks per rotation, or just one.
I think the filawinder and the feed knocker will remove a lot of variation and make it hands off.
The only other thing I think I would do if I would get two power sources and move the heater and motor off of the same set-up. I can definitely hear the effect of the heater on the motor. I don't know what the effect is on the filament and it definitely affect the cost. The motor doesn't have to be connected to the controller, does it? It could just be on a separate power supply with a switch?
Between those additions (not even the dual controller)
Either that or finding a 12V motor with an offset/unbalanced shaft and wiring that into the motor circuit to act like a constant buzz.
Distinct, larger taps or a constant buzz- What do you think will work better.
One thing I've found is that just because the extruder is up to temp doesn't mean that the plastic is ready to be extruded. It takes my extruder about 12 min to get to 180C. I need to let it settle for about 10 more minutes so that the plastic is melted and the extruder doesn't try to twist itself apart or slip the power coupling. I stop and start it a few times rather than just turn it on. That sound Kosher for ABS?
Just out of curiosity, how do filament manufacturers get the filament to wind so perfectly on the spools? If they are doing it by hand, I understand why filament is so expensive!